Axcelis Technologies (NASDAQ:ACLS) Q1 2026 Earnings Call

0:00 Only ask one question and one follow-up. Please stand by while we compile the Q and A roster. Our first question comes from the line of Jed Dorsheimer of William Blair. Your line is now open. Hey, thanks guys. I was just wondering, Russell, maybe you could I know the gallium nitride and indium phosphide are mostly deposition which makes your merger that much more exciting but there is some implant associated with containment etcetera. So I was wondering if you might just spend a minute just talking about that opportunity. And then I have a follow-up. Thanks.

0:46 Yes. Hi, Jed. So are you talking about the implant opportunity in silicon photonics, particularly in data centers? That where you're heading? Yes. Well, yeah, that's where I'm heading, but, you know, specifically around indium phosphide and and then gallium nitride on the power side. Sure. So there's not really many, but there are a couple of implants in gallium nitride, but they're not significant. What I would say when you think about data center and optical transmission, we don't partake, especially, in the laser manufacturing part. But to to put the data, to encode the laser with the data, that requires a modulation unit. And that modulation unit is a silicon unit, and that does require implantation.

1:33 I think it's essentially an isolation implant that is doing. So again, it's not a massive use of ion implantation, but it's definitely an application in certain photonics for that particular area. Got it. Thanks. So it'll mostly be on the Vico product line post merger, correct? Yes, I think that's true. I think they have I think you're on their call. They have the MOCVD and the thin film capability for the optical components. Thank you. I'll jump back in queue. Thanks. One moment for our next question. Our next question comes from the line of Dennis Pichatzin at Needham and Company. Your line is now open.

2:24 Great. Thank you very much. So, the outlook, it sounds like you're seeing some strength in the power site C and memory, but general mature remains weaker, so it may be getting even weaker. Can Can you discuss in some more detail what you're seeing in general mature? I mean, it seems like this quarter that some IDM business picking up globally for some of the mature products, but doesn't seem like you're seeing any of that. Can you talk a little bit more please about what you're seeing in general mature? Yes, Dennis, this is David. So I think we would agree that the end markets are picking up. I think oil is stabilizing, industrial recovering, and we're starting to see AI pick up for this general mature segment. When we talk about the Q1 bookings, yes, we did see a little bit softer bookings in general mature, but the key takeaway there is that utilization rates are rising.

3:16 And so for us, that's a really nice data point. We still expect 2026 general mature to be down year over year. But as we think about the business, trending into 2027, we feel a little bit better today than we did three months ago in general. And on the silicon carbide side, in Russell's prepared remarks, quite clear that we are seeing also relative to three months ago, I would say encouraging signals from our customers. Conversations are very constructive. There's a lot of inbound about technology, engagement, channeling, superjunction, and then you're starting to hear more and more about silicon carbide in the data center.

4:03 So we think the combination of these things is pretty encouraging and again, sets the stage for maybe some momentum into 2027. And from quarter to quarter, it can move around. But we're taking a bigger picture view that we think that next year looks a little better for these markets. Great. And then for my follow-up, so on a related note, the book to bill for the quarter was up to about one, I think up from 0.8. Could you share a little more about what you expect to see order wise next quarter? Is the book to bill going to be the same or moving even higher? Yes, we're not going to guide bookings per quarter. It can fluctuate. But we are encouraged that Q1 was relatively consistent with Q4, which was a big step up from earlier in 2025.

4:56 So this is a good sign, but we're not going to be forecasting bookings on a quarterly basis. Understood. Well, that's it for me. Thank you very much. Thanks, Yut. Our next question comes from the line of David Dooley of Steelhead Securities. Your line is now open. Good afternoon, and thanks for taking my questions. I guess I'll ask the question I think I've asked in the last couple of conference calls. If you could elaborate on what you're seeing as far as silicon carbide adoption in other markets, specifically in the data center. I think in the past you've talked about voltage step down applications. I was just wondering if you're getting closer to the rack or if there's other applications that you might be addressing inside the data center realm.

5:57 Hey, Dave. It's Russell. Thanks for the question. So, yeah, obviously, data center is an application for it. I just wanna start off with the electric vehicle. I think they're actually very exciting right now. A lot of the cars are going to 800 volt subsystems, which allows them to actually charge so much quicker as well. So you're seeing that trend and those are all using silicon carbide. The first application would be the traction inverter. And you're seeing more and more of the silicon carbide cars now taking silicon sorry, of the electric vehicles now taking silicon carbide. But we're also seeing more and more silicon carbide per EV. So you have the traction inverter, which is typically the first device they take. Then you start to see the DC to DC step downs, also the onboard charging.

6:43 And of recent times, we're seeing the AC system, the compressor having silicon carbide. So we're really seeing this next generation of cars much greater range and much faster charge types. I think those are that's really good trend. And although it may be a little bit soft now, the long term trend is that's going to continue to grow. So that's kind of the electric vehicles. Like I mentioned before I get on to data centers, this has actually it surprised me when I when we're in China and we were told about these big and white goods. That that really was surprising. But, yeah, for for many of these high end systems that are running about a kilowatt, they need this to achieve their certification for for for power efficiency in multiple markets. And then the data center, it was interesting. So maybe three to six months ago, we'd be talking about, is it going to be gallium nitride?

7:33 Is it going to be silicon carbide? I think at this stage, it's becoming clearer that it's both. As the rack transitions to 800 volts, I think what you're going to see is gallium nitride inside the rack. That's a great sweet spot for gallium nitride. But as you drop down from the grid, which 13 kilovolts or more, that's going to be a great application of silicon carbide. So as it drops from the grid down towards the rack, you're going to see silicon carbide taking that opportunity as it really is a good application of that material. And as a follow-up or just an extension of that question, do you think this 800 volt rollout in the data centers, which I guess kind of starts in the middle of this year, really gets going next year. Is that when you might see more of a pickup in your silicon carbide implant business for that market? Or when will this start to hit a more significant way, I guess, is the best way to ask it?

8:34 Dave, that's a good question. It's still low volume relative to EVs. But I would say our conversations with customers, they're talking a little bit more about it. And so we're paying attention. And this could be an application that over time could grow, whether it's the transformers or other parts of the data center power architecture. So I think TBD, Dave, how big this could be. But I'd say our customers are definitely paying a lot more attention to that. All right. That's good news. Thank you. Thanks. One moment for our next question. Our next question comes from the line of Zhu Kun Chang of Bank of America Securities.

9:24 Your line is now open. Hi, thank you for taking the question. One on the full year 2026 guide. I think you said you want to you're maintaining that flattish year over year guide. I'm just curious because clearly memory sentiment and new wafer starts are likely much better than ninety days ago. And then at the same time, I think you said your bookings for the general mature and the power business are still doing pretty well. So I'm curious why not raise that full year guide? Thank you. Yeah, so we feel comfortable with flat year over year. Absolutely, we expect to be a strong year for us in 2026, offset by a digestion and power in general mature. So we're seeing encouraging signs.

10:18 We like what we saw in the bookings in Q1. One quarter is not a trend. But I think that some of that momentum is probably going to carry into 2027. For this year, based on our backlog, based on the scheduled shipments that we have, we still feel that those markets are going to be down this year. And so net net, you're going to get we expect about flat, but with a nice setup into 2027. Got it. And then a follow-up more specifically on the memory side. So I think at one point you said memory is going to be up year over year and clearly the trends are looking that way. How should we think about the full year trajectory when obviously it tends to be a little bit lumpy on a quarterly basis, but we had a pretty strong start to the year.

11:11 And I think also NAND activities have begun to pick up. So any color on the full year memory would be helpful. Thank you. Yes, that's right. From quarter to quarter, it could be lumpy. It just depends on when customers are taken when we're shipping to customers and fab availability space. So for example, in Q2, memory is going be a little softer sequentially. But we think for the full year, it's we expect pretty strong growth in memory. And that's all DRAM. Our expectation embedded in our twenty twenty six expectation is not much NAND. So that's something that we're paying attention to. As you know, Duxin, a lot of the NAND manufacturers are focused on vertical scaling.

12:02 And so that doesn't require a lot of incremental implant. But once the industry starts to add wafers, that's when we would start to see a pickup. Got it. And then if I may, just one quick one. Within memory, and sorry for being too specific maybe, but I think the industry is trending and shifting away a little bit from HBM and diversifying more into other types of memory. And I assume wafer consumption ratio is not as high on those, for example, like the LTDRs. Obviously, I think that would be a positive trend for you. Have you seen any signs that the demand signals are stronger because of that?

12:53 I don't think we've seen any meaningful change based on any mix shift in memory. So customers are still really trying to build out HBM capacity and DRAM for AI. Yes. And when I think about HBM, it's essentially DRAM stacked. And so the implant intensity doesn't change much between DRAM and HBM. And NAND is not different. It's a slightly different mix of high energy, high current, medium current. But I'd say that NAND and even node to node to node as well, it's not changing significantly. I'd say the implant intensity is relatively stable. All right. Thank you so much. Thank you. I'm showing no further questions at this time. I would now like to turn it back to David Rizik for closing remarks.

13:42 Thank you, operator. Just wanted to thank everybody for joining the call and your interest in Axcelis. You can close the call. This concludes the presentation. Thank you for your participation in today's conference. You may now disconnect. Have a good day.